Growth of the transplantable EMR-86 rat mammary carcinoma depends on elevated prolactin levels which are induced by oestrogenic stimulation of the pituitary. We investigated histological and cell kinetic changes during tumour regression after removal of implanted oestrogen pellets (EP), and we especially focused on the role of apoptosis. After EP removal, serum prolactin decreased to basal levels in 5 days, reaching its largest depletion during the first day. Similarly, S-phase cell fractions, assessed by bromodeoxyuridine (BrdUrd) incorporation, decreased to half the initial value during the first day and developed into a gradual decrease to basal levels thereafter. Within 10 days, tumour volumes were reduced to 20% without striking changes in tissue architecture. To quantify apoptosis, we applied a method that stains DNA breaks in tissue sections and subsequently measured the stained area by automated image cytometry. This procedure was necessary, as the subtle changes could not be detected by histological examination alone. One day after the rapid decline of the S-phase fraction, a 3-fold increase in apoptotic area was observed that remained for about 3 days and then gradually decreased. This correlated with the histologically observed reduction of tumour cells. In spite of the major cell loss, regression came to a halt after about 10 days. The surviving cell fraction is discussed within the context of a stem cell hypothesis, in which tumour cells with stem cell characteristics are less susceptible to hormone-induced apoptosis than their (non-stem) daughter cells. This notion has implications for the eradication of residual tumour cells, because a diminished susceptibility might also apply to apoptosis induced by radio- or chemotherapy.